2026-07-21 06:05 UTC
@hyc@mastodon.social @w3cdevs@w3c.social @w3c@w3c.social Hi Howard, normalization is only one component of string matching. The scope of the issues addressed by this document is far broader.
For instance, normalization ignores case. Whether "Green" and "GREEN" should be considered a match depends entirely on the context.
There are also issues regarding character escaping, legacy encoding conversions, invisible characters, and more.
Replies (1)
-
@xfq@w3c.social 2026-07-21 06:05
@hyc@mastodon.social @w3cdevs@w3c.social @w3c@w3c.social In fact, the document’s core recommendation is *not* to perform normalization. Normalization is introduced only as an internal algorithmic step in specific scenarios, as it risks destroying distinctions that the user intentionally preserved. Even with normalization, strings that appear identical may still fail to match.